The project ?Metabolomics of longevity? will partner with all projects and cores within the Longevity Consortium study. Its objective is to detail the metabolic biomarkers and biochemical mechanisms that differ betwee long- lived species or subjects, in comparison to species or subjects with shorter lifespans. We specifically investigate balancing nutritional energy, repair and prevention of metabolic damage, and metabolic responses to stress. We further hypothesize that sustaining a healthy metabolism correlates with specific molecular signatures that are acquired in other projects of the Longevity Consortium, so that metabolomics data can be integrated into context analysis in a meaningful manner. Consistent with the overall emphasis and design of the Longevity Consortium, we will focus on pathways relevant to control of metabolic regulatory capacity rather than on disease-specific signatures. According to these overarching aims, we will acquire and interpret high-quality metabolite data by using targeted and untargeted mass spectrometry. We will identify and quantify over 800 known metabolites, in addition of over 2,000 metabolic signals that lack structure identification. Metabolite classes will cover primary amines, bile acids, steroids, inflammatory oxylipins, complex lipids, biogenic amines and miscellaneous compounds, including dietary and drug exposome markers. Specifically, in aim 1, we will randomize samples from four large human cohorts to compare baseline metabolomics markers of 683 subjects who continued to live longer than 98% of subjects of the U.S. population, and compare these to 2,049 subjects who had a shorter life span. We will integrate several statistical tools to compile a panel of longevity biomarkers that will be validated by an independent cohort of 450 subjects who lived longer than 100 years. We will investigate all data in a longitudinal manner by appropriate statistical tools to predict changes in a range of age-related phenotypes such as grip strength and walking speed that ultimately may contribute to longevity. In aim 2 we will analyze cells from over 100 long-lived and short-lived species to achieve mechanistic understanding of conserved metabolic differences in cellular metabolic homeostasis. We will also compare long-lived mice against shorter-lived wild type mice, including through lifespan-extending drugs such as rapamycin and acarbose. In aim 3, we will combine literature-based food- and microbial metabolic markers into the analysis of the human cohort studies as confounding factors that . may contribute to human longevity and that are not under human genetic control. We will also contribute to efforts in the Orwoll-proteomics, the Girke- chemoinformatics and the Price-systems cores by providing comprehensive metabolite/gene and metabolite/protein annotations through integrating existing biochemical databases. We will supplement these analyses by phenotypic context information for metabolite data through text-mining association analyses.